| KEYWORD |
Data science for software engineering in EIS
Thesis in external company
keywords DATA ANALYTICS, SOFTWARE ENGINEERING PROCESS
Reference persons MAURIZIO MORISIO
Research Groups GR-16 - SOFTWARE ENGINEERING GROUP - SOFTENG
Thesis type EXPERIMENTAL AND MODELING
Description The thesis is about the analysis of data generated by the software production process in a mainstream software development company. The available data is about commits, releases, defects and their causes, requirements, human resources, iterations, effort and cost.
The goal is twofold:
-develop models to explain and predict defectivity (what kind of defects are inserted, where and why, what are their root causes)
-develop models to explain and predict development time and effort (which factors explain and predict effort, duration, iteration of software projects)
Step 1 is about developing connectors to collect raw data from a variety of development tools and platforms used in a variety of projects
Step 2 is developing data conceptual models to aggregate and analyse data
Step3 is about applying different categories of techniques (statistics, machine learning, deep learning) to build models capable of explaining and predicting relationships in the data
Stap4 is about testing the models on past and current projects from the company
Given the amount of work needed the work could be split among more students, at least one for defectivity, one for effort explanation
Required skills software engineering, data analysis techniques, Python, R, Java
Deadline 24/04/2020
PROPONI LA TUA CANDIDATURA